feat: 新增 code_stats Skill，修复 Skill handler 参数解析

- 新增 skills/code_stats.py: Python Skill 示例，统计代码量
- 修复 unified_tool_manager: Skill handler 支持 key=value 参数格式
- Skill 协议测试通过: 直接调用 + MCP 调用均正常

本次更新由 mimo 自动化执行，与开发者无关。

skills/code_stats.py

@@ -0,0 +1,140 @@
+"""
+代码统计 Skill - 统计项目代码行数、文件数等
+"""
+import os
+from skills.base import Skill, SkillResult, SkillParameter, ParameterType
+
+
+class CodeStatsSkill(Skill):
+    """代码统计技能 - 统计目录中的代码行数、文件数、注释率等"""
+
+    def __init__(self):
+        super().__init__()
+        self.name = "code_stats"
+        self.description = "统计项目代码量：文件数、代码行、注释行、空行、注释率"
+        self.version = "1.0.0"
+        self.author = "mimo"
+        self.tags = ["代码", "统计", "stats", "lines"]
+
+    @property
+    def parameters(self):
+        return [
+            SkillParameter(
+                name="directory",
+                param_type=ParameterType.STRING,
+                description="要统计的目录路径",
+                required=True
+            ),
+            SkillParameter(
+                name="file_pattern",
+                param_type=ParameterType.STRING,
+                description="文件匹配模式，如 *.py",
+                required=False,
+                default="*.py"
+            ),
+        ]
+
+    def execute(self, **kwargs) -> SkillResult:
+        directory = kwargs.get("directory", ".")
+        file_pattern = kwargs.get("file_pattern", "*.py")
+
+        if not os.path.isdir(directory):
+            return SkillResult(success=False, error=f"目录不存在: {directory}")
+
+        # 确定文件扩展名
+        if file_pattern.startswith("*."):
+            ext = file_pattern[1:]
+            file_filter = lambda f: f.endswith(ext)
+        else:
+            file_filter = lambda f: True
+
+        total_files = 0
+        total_lines = 0
+        total_code = 0
+        total_blank = 0
+        total_comment = 0
+        file_details = []
+
+        skip_dirs = {'.git', 'node_modules', '__pycache__', '.venv', 'venv', 'env'}
+
+        for root, dirs, files in os.walk(directory):
+            dirs[:] = [d for d in dirs if d not in skip_dirs and not d.startswith('.')]
+
+            for filename in sorted(files):
+                if not file_filter(filename):
+                    continue
+
+                filepath = os.path.join(root, filename)
+                rel_path = os.path.relpath(filepath, directory)
+                total_files += 1
+
+                file_lines = 0
+                file_code = 0
+                file_blank = 0
+                file_comment = 0
+
+                try:
+                    with open(filepath, 'r', encoding='utf-8') as f:
+                        for line in f:
+                            total_lines += 1
+                            file_lines += 1
+                            stripped = line.strip()
+                            if not stripped:
+                                total_blank += 1
+                                file_blank += 1
+                            elif stripped.startswith('#'):
+                                total_comment += 1
+                                file_comment += 1
+                            else:
+                                total_code += 1
+                                file_code += 1
+                except (UnicodeDecodeError, PermissionError):
+                    continue
+
+                file_details.append({
+                    "file": rel_path,
+                    "lines": file_lines,
+                    "code": file_code,
+                    "comment": file_comment,
+                    "blank": file_blank,
+                })
+
+        # 按代码行数排序
+        file_details.sort(key=lambda x: x["code"], reverse=True)
+
+        # 构建报告
+        comment_rate = (total_comment / max(total_code, 1)) * 100
+
+        report_lines = [
+            f"📊 代码统计 ({directory})",
+            f"{'─' * 40}",
+            f"  文件数: {total_files}",
+            f"  总行数: {total_lines}",
+            f"  代码行: {total_code}",
+            f"  注释行: {total_comment}",
+            f"  空行:   {total_blank}",
+            f"  注释率: {comment_rate:.1f}%",
+        ]
+
+        if file_details:
+            report_lines.append(f"\n📄 文件明细 (按代码行数排序):")
+            for fd in file_details[:20]:
+                report_lines.append(
+                    f"  {fd['file']:40} {fd['code']:5} 行代码"
+                )
+            if len(file_details) > 20:
+                report_lines.append(f"  ... 还有 {len(file_details) - 20} 个文件")
+
+        return SkillResult(
+            success=True,
+            message='\n'.join(report_lines),
+            data={
+                "total_files": total_files,
+                "total_lines": total_lines,
+                "total_code": total_code,
+                "total_comment": total_comment,
+                "total_blank": total_blank,
+                "comment_rate": comment_rate,
+                "files": file_details
+            }
+        )

unified_tool_manager.py

@@ -206,11 +206,20 @@ class UnifiedToolManager:
             kwargs = {}
             
             # 解析参数
-            if args.strip().startswith('{'):
+            args = args.strip()
+            if args.startswith('{'):
                 try:
                     kwargs = json.loads(args)
                 except json.JSONDecodeError:
                     kwargs = {'operation': args}
+            elif '=' in args:
+                # key=value 格式: "directory=. file_pattern=*.py"
+                for part in args.split():
+                    if '=' in part:
+                        k, v = part.split('=', 1)
+                        kwargs[k.strip()] = v.strip()
+                    else:
+                        kwargs.setdefault('operation', part)
             else:
                 kwargs = {'operation': args}